AutoSkill: Experience-Driven Lifelong Learning via Skill Self-Evolution

TL;DR

AutoSkill is a framework that enables large language models to autonomously learn, evolve, and reuse personalized skills from user interactions, enhancing lifelong personalization without retraining.

Contribution

It introduces a model-agnostic plugin system for extracting, maintaining, and dynamically injecting skills from interaction traces, facilitating lifelong learning in LLMs.

Findings

AutoSkill effectively derives skills from dialogue traces.

It supports continual self-evolution of skills.

AutoSkill enables personalized capabilities to be reused across sessions.

Abstract

In practical LLM applications, users repeatedly express stable preferences and requirements, such as reducing hallucinations, following institutional writing conventions, or avoiding overly technical wording, yet such interaction experience is seldom consolidated into reusable knowledge. Consequently, LLM agents often fail to accumulate personalized capabilities across sessions. We present AutoSkill, an experience-driven lifelong learning framework that enables LLM agents to automatically derive, maintain, and reuse skills from dialogue and interaction traces. AutoSkill abstracts skills from user experience, supports their continual self-evolution, and dynamically injects relevant skills into future requests without retraining the underlying model. Designed as a model-agnostic plugin layer, it is compatible with existing LLMs and introduces a standardized skill representation for sharing and transfer across agents, users, and tasks. In this way, AutoSkill turns ephemeral interaction experience into explicit, reusable, and composable capabilities. This paper describes the motivation, architecture, skill lifecycle, and implementation of AutoSkill, and positions it with respect to prior work on memory, retrieval, personalization, and agentic systems. AutoSkill highlights a practical and scalable path toward lifelong personalized agents and personal digital surrogates.